No, Actually, the Fiscal Gap Has Not Fallen by 8 Percentage Points

Last week, at a breakfast with David Wessel and Gerald Seib of The Wall Street Journal, Council of Economic Advisers chair Jason Furman made the case that not only had short-term deficits been coming down, but also that the long-term situation has improved dramatically. Specifically, Furman suggested that the 75-year fiscal gap has fallen from almost 10 percent of GDP to under 2 percent.

While improvements have occurred in recent projections, however, our analysis suggest these improvements are far smaller than what Furman suggested. The true improvement in the fiscal gap is far closer to 0.8 percent of GDP than to 8.0 percent. Furman's comparisons are largely apples and oranges.

In his talk, Furman said the following (emphasis added):

[CBO's] fiscal gap over 75 years is a little bit less than 2 percent of GDP. Now, if you look at the debt 75 years from now it spirals and goes really high but every time you have a small wedge things get very large. In contrast, I used to write things before CBO was doing the estimate with Alan Auerbach and Bill Gale and we generally had numbers that were nearly 10 percent of GDP. So the 75-year outlook is considerably better and that’s because of projections around health and some of the changes we’ve made in terms of revenue and things like the Affordable Care Act that has significant long-run deficit reductions.

Furman is correct that the fiscal gap -- the amount of non-interest deficit reduction needed each year (starting immediately) to maintain the current level of debt 75 years from now -- in CBO's current law baseline is less than 2 percent of GDP. To be exact, CBO estimates a 75-year fiscal gap of 1.7 percent of GDP (in our Realistic baseline, we estimate the gap to be somewhat higher at 2.7 percent). But the suggestion that it has come down by something like 8 percent of GDP is quite misleading for at least four reasons:

  1. Auerbach, Furman, and Gale (hereafter AFG) never estimated a 10 percent fiscal gap. Their 75-year fiscal gap estimates in 2007 and 2008 ranged from 2.9 to 7.3 percent of GDP, depending on the assumptions.
  2. The higher AFG estimates made current policy baseline adjustments, whereas CBO's fiscal gap analysis is based on strict current law.
  3. The AFG estimates assume long-term revenue and non-health, non-Social Security mandatory spending are frozen as a share of GDP after ten years, whereas CBO assumes they grow and shrink, respectively. Both of those AFG assumptions significantly worsen the long-term outlook.
  4. Debt is more than twice as high as when the AFG estimates were made, and since the "fiscal gap" measures the necessary deficit reduction to stabilize the debt at current levels, the size of the fiscal gap actually actually shrinks when current debt levels grow.

When you account for these factors, it is clear that the long-term fiscal gap has not fallen by anywhere near 8 percentage points. Even comparing the strict numbers from AFG in 2007, one could only claim a reduction of a bit more than half that amount a 4.6 point reduction from 6.3 percent of GDP down to 1.7 percent.

Yet even that reduction is overstated. The AFG baseline assumed various tax cuts were extended, the wars drew down, and discretionary spending grew faster than inflation, which if assumed today would increase CBO's 1.7 percent gap to more like 2.2 percent.

Much more significantly, AFG assumed that after the end of the decade, revenue and mandatory spending would be frozen as a share of GDP. This is in serious contrast to the CBO baseline, which projects revenue will rise by 6 percentage points of GDP between 2023 and 2088 and other mandatory spending will fall by 1.5 percentage points. This difference is responsible for about 3 points in the fiscal gap.

On top of these adjustments, if we set the goal of achieving 2007 debt levels in 2088, rather than 2013 debt levels, this would further increase the gap. All told, closing this gap would require immediate deficit reduction totaling 5.5 percent of GDP, only 0.8 points less than the 6.3 calculated by Auerbach, Furman, and Gale in 2007.

Bridge from CBO Current Law to AFG Assumptions (Percent of GDP)
Assumption 75-Year Present Value
CBO Current Law 1.7%
CRFB Realistic Ten-Year Adjustments +0.2%
Extend Normal Tax Extenders +0.3%
Freeze Other Mandatory Spending and Revenue after 2023 +2.9%
Subtotal, 75-Year Fiscal Gap 5.1%
Set Debt Target to 2007 Levels +0.4%
Total, 75-Year Fiscal Gap to Achieve 2007 Debt Levels
5.5%
   
Memorandum: 2007 AFG Fiscal Gap 6.3%

 Source: CBO, CRFB calculations

To be sure, our 5.5 percent estimate is not a perfect apples-to-apples comparison either. One could argue that we should only calculate the gap through 2081 as AFG originally did, for example, which would reduce the fiscal gap by 0.4 percent. Or one could argue that we should assume automatic cuts from sequestration remain in place through 2021 and continue beyond their expiration, which would reduce the gap by another 0.4 percent. Yet even combining these assumptions with an assumption that all of the tax extenders and SGR cuts will be fully offset on a same-year basis would yield a fiscal gap between 4 and 4.5 percent of GDP.

The major difference between CBO and AFG is not the policies that have been enacted or even the policies that are assumed, it is the long-term assumptions on revenue and non-health/retirement mandatory spending. Whether those categories are frozen (as in AFG) or not (as in CBO) accounts for a 3 point difference in the fiscal gap. Perhaps CBO's assumptions are too generous, or perhaps the AFG assumptions were too conservative (or perhaps both), but either way, comparing them on a like basis is a clear case of comparing apples and oranges.

So if AFG and CBO are apples to oranges, what about CBO and CBO? Unfortunately, even CBO assumptions change from year-to-year, and it is unclear when that change is truly a result of new policy versus new thinking. But if one were to compare CBO baselines over time, they’d be hard pressed to find a major improvement. Since 2007, CBO has calculated the 75-year fiscal gap every year (except for 2008) using its extended baseline scenario and its alternative fiscal scenario. Although estimated have bounced around significantly, they are nearly identical today to what they were in 2007. Under CBO’s current law scenario, the fiscal gap increased from 1.7 percent in 2007 to 3.2 percent in 2009 before falling to -1.1 percent by 2012 and then returning to 1.7 this year. Similarly in the AFS, the gap went from 6.9 percent in 2007 to 8.7 percent by 2010 to 7.1 percent this year.

Source: CBO, CRFB estimate for 2013 AFS

Furman uses CBO numbers to make the case that "the long run deficit outlook is also considerably better than it used to be." Although there have certainly been improvements, both as a result of enacted legislation and a slowdown in health care cost growth, we wouldn’t call the long-term improvements "considerable." CBO’s own budget director, Doug Elmendorf, agrees. As he explained only a few months ago (emphasis added):

The federal budget deficit has fallen faster than we expected. However, relative to the size of the economy, debt remains historically high and is on an upward trajectory in the second half of the coming decade. The fundamental federal budgetary challenge has hardly been addressed: The largest federal programs are becoming much more expensive because of the retirement of the baby boomers and the rising costs of health care, so we need to cut back on those programs, increase tax revenue to pay for them, or take some combination of those actions.

Policymakers must address our long-term fiscal challenges, not simply assume the problem away.

Fiscal Gap Measures

Nice job of reconciling disparate estimates. The CBO to CBO comparisons were very useful.

Post a New Comment

  • Web page addresses and e-mail addresses turn into links automatically.
  • Allowed HTML tags: <a> <em> <strong> <cite> <code> <ul> <ol> <li> <p> <br><img><div><span><object><embed><blockquote> <!--break-->
  • Lines and paragraphs break automatically.
  • Insert a chart by placing [chart:nid] into your content, where nid is the node ID.

More information about formatting options

By submitting this form, you accept the Mollom privacy policy.